1 Pull from the course Github

git pull origin master

2 First, a not-so-shiny story

I came to SIO in Dec 2012 with the task of writing a Graphical User Interface (GUI) to run stable isotope mixing models. I ended up using an R package called gwidgetsRGtk2, which did the job:

library(MixSIAR)
mixsiar_gui()

2.1 Why didn’t I use Shiny?

  • It was new and not well documented in 2013. This has HUGELY changed!
  • Didn’t appreciate the momentum behind R Studio
  • The “server” talk made me think it had to be hosted online (not true, can be deployed locally)
  • Seemed designed for lightweight computing, not long model runs

I am still not sure if it would have been more or less straightforward to use Shiny instead of gWidgetsRGtk2… but if I were starting from scratch I would use Shiny. Mostly based on the user community and documentation/examples available now.

3 1. “Get inspired.” What is a “Shiny app”?

Movie explorer

Health outcomes movie

3.1 It’s got “widgets”

Widget gallery

3.2 When would I use a Shiny GUI/app?

  • Expand reach of your research by reducing barriers to others using your

  • Teaching courses (ex: regression)

  • Data exploration

  • Data analysis (ex: classify plankton)

4 2. “Get started” (tutorial)

Because I have never used Shiny, let’s let R Studio explain the basics:

https://vimeo.com/rstudioinc/review/131218530/212d8a5a7a/#t=4m08s

4.1 Lesson 1

https://shiny.rstudio.com/tutorial/lesson1/

library(shiny)
runApp("/home/brian/Documents/SIO_misc/SIO-R-Users/Intro-Data-Viz-Winter-2017/week-9/01_hello")

5 Resources